Harmonics and VFDs 22

[Marke]

One of the areas that I operate in, has a high density of VFDs on pumps on relatively weak supplies. The result is that the high levels of harmonics on the VFD inputs has accumulate in the supply and is causing a high THD of the supply voltage. While we do have supply regulations covering harmonics, in this instance, the harmonics are higher than they should be.

There is an option of using zig zag transformers and six phase rectifiers as a means of reducing the harmonics drawn by drives however in this case, the drives are already installed.

There is a transformer for each drive and sizes range up to 200KW.

One thought that I had, was that for future installations, and there are new installations going in all the time, that the new supply transformers be designed with a zig or a zag winding to give a phase shift, and install equal loading on the leading and lagging phase shifts. This should act like a twelve pulse input on one drive, only it will be across two drives.

Any thoughts on this??

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[itsmoked]

Interesting point LionelHutz.

I wonder if a really aggressive negotiation with them for several hundred might not get the price down substantially. They could also then prep them so they are a pull-out-of-the-box-screw-in-place-wire-in-series-close-lid kind of convenience.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[dpc]

The individual transformers are effective at reducing harmonics. The down side is that both motor have to be running at similar loads to get the biggest benefit. It won't help at all if only one motor is running, so it certainly is not as widely applicable as 12-pulse drives.

If individual isolation transformers are being used anyway, there isn't much downside to buying transformers with different phase shifts.

But I do agree that a good filter can also be effective. I'm familiar with Mirus. Their broadband filter seems to have some good features and they are fairly popular in some industries.

 

[Valvecrazy]

Of course a VFD works better with a pump that has a steep curve. This usually means that you have to oversize the pump to be able to (save energy?) by slowing it down with a drive. With the pump you picked the VFD will vary the output from 100 HP at 1200 GPM to 38 HP at 100 GPM. If you add back in the parasitic losses and the loss of motor efficiency from running on a VFD, the load on this pump will actually be about 103 HP at max flow and 40 HP at low flow.

Attached are two more curves for Submersible pumps.

With the curves I picked, the SUBMERSIBLE pump only uses about 85 HP to pump 1200 GPM. Restricting the flow with a valve on these pumps will still reduce the required horse power from 85 HP to about 45 HP.

The pump you picked for the VFD will use 18 HP more at max flow, than the pump I picked to work with a valve.

The pump you picked for the VFD will only save about 5 HP at minimum flow over the pump I picked to work with a valve.

Therefore, if the pump is used the same amount of time at high flow as it is at low flow, the VFD will cause the pump to use an average of 10 HP more power.

The way so called efficiency of the pump is figured for the VFD, is deceiving and makes a VFD look better than it really is. The curve being flat or steep has much less to do with anything than how fast the electric meter is actually spinning.

Submersibles are about 80% of my business, and if the head requirement is 230', it will always be 230'. The well is not going to change from 230' deep to 120' deep just to help a VFD be more efficient. When the head is constant, as it is with nearly all pumping applications, it doesn't matter if it is submersible, centrifugal, or turbine, I can nearly always pick a pump that will save just as much energy by using a valve, as when using a VFD.

Look at it this way. When the pump has been up-sized to work with a VFD, high flow of 1200 GPM at 100 HP is using .0833 HP per gallon. This is not even counting parasitic and other losses from the VFD. When slowed with a VFD to 100 GPM and using 40 HP, that is .4 HP per gallon produced. That means when the VFD is being used for low flow, it is actually burning .3167 extra horse power per gallon produced.

When the pump has been picked properly, it only uses 85 HP to pump 1200 GPM, which is only .0708 HP per gallon produced. That means that the VFD is burning .0125 HP at high flow when compared to a properly sized pump running Across The Line and using a valve for control.

The valve I am referring to only has 7 PSI friction loss at 1200 GPM, which is the only time a valve is actually burning energy. When a valve is used to further reduce the flow, the excess head produced is actually a free by-product, as throttling causes a reduction in power.

In that respect, a VFD always burns just as much energy, and in many cases more than a valve.

Anytime you can do the job without using a VFD, there are a multitude of problems, which harmonics are just one of, that are never present to start with.

 

[gepman]

valvecrazy

You should put a post in either the pump engineering (under Mechanical) or energy engineering general (under Energy) if you are this passionate about the subject. I see how you came up with your handle!

I believe that you have some valid points which you have carried to an extreme. I would be happy to point out what I believe are the errors in your logic in another forum.

 

[Marke]

Hello valvecrazy

With the pump you picked the VFD will vary the output from 100 HP at 1200 GPM to 38 HP at 100 GPM.

I am curious as to how you got this figure. Can you please enlighten me.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[Valvecrazy]

Thanks gepman! I studied electronics, including VFD's in school. I sold and serviced VFD equipment until about 1993, and became increasingly frustrated with all the problems with VFD controls. I was using valves as a bypass for the VFD's, and observed that the power consumption was basically the same when the valves where being used, as when the Drives were actually working. I then started pulling out any VFD I had installed and replaced them with across the line panels and control valves. With valves, my installations became more reliable, less expensive, just as efficient, and nearly all my problems went away. Since then I have installed many thousands of valve controlled systems. I would be more than happy for you to point out any errors in my logic. I have been accused of many things for my stance on this subject. I believe the only thing I am guilty of, is knowing how to actually read and understand a pump curve. We may should start another thread however, the question was how to solve harmonic problems with VFD's. I believe I have the best solution.

Marke, I used the 100 HP and 1200 GPM right off your pump curve. For the 38 HP at 100 GPM I used the affinity law. For your pump to be producing anything at 230' of head, it has to still be spinning at 72% of full speed. Because head drops off by the square of the speed, you will only have 50% of the shut off head at 72% speed, which gives me 230'. Because HP drops off by the cube of the speed, the pump is still pulling 38 HP at the minimum possible speed of 72%.

 

[gepman]

Marke, I presently consult on energy efficiency to the two largest electric utilities in California and haven't heard of this being a problem (although I am just in the process of obtaining a power quality analyzer) but I doubt that even in our rural areas we are quite as rural as in NZ. The utilities here believe in VFD's and give (cash) incentives to put them in, of course the application has to be correct. Please post any solution which works because I am sure that with the increasing use of VFD's in California's rural areas that this could become a problem.

valvecrazy

Cut and paste your posts into a new thread in one of the two suggested forums that I gave and I will respond point by point.

 

[Marke]

Hello valvecrazy

If you insert a valve in the line to reduce the flow, what happens to the pressure on the output side of the valve?
i.e. the head across the load.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[Valvecrazy]

"The valve I am referring to only has 7 PSI friction loss at 1200 GPM, which is the only time a valve is actually burning energy. When a valve is used to further reduce the flow, the excess head produced is actually a free by-product, as throttling causes a reduction in power."

In other words, there is 7 PSI of friction loss across the valve at maximum flow rate. The output of the valve stays at a "constant pressure", say 50 PSI regardless of the flow rate. The pump must produce max flow rate at 57 PSI to accomplish this. This is as long as the valve stays loaded and ready. I have several ways of dumping the water from the control chamber of the valve, which reduces the friction loss across the valve to about 1 PSI. This is rarely needed as the 7 PSI is usually not a problem. I always get quite a bit more pressure and flow from the pump when it is running ATL than when it is controlled by a drive. I find that the loss of motor efficiency and the parasitic losses of the Drive itself, cause more of a loss on pump performance than the 7 PSI friction loss across the valve.

The whole idea of choking a pump back to conserve energy is "counter intuitive". Our minds have no trouble understanding that slowing a pump down reduces energy consumption. The fact that power is reduced almost exactly the same by simply restricting flow with a valve, is very difficult to grasp. Most people make the false assumption that throttling waste or burns energy. I have several working demonstrators that can easily be switched from VFD to ATL and valve control. Every pump engineer that has seen this demonstration, is bewildered and you can actually see the bulb light up over their head, when they see for themselves that throttling with a valve reduces energy consumption just as much as a VFD.

I am also working with one of the largest utilities in California. At least one of the main guys there told me that he understands that throttling can reduce energy consumption of a pump as much as a VFD. As he said, the soft start may be the only advantage of a VFD, which could easily be accomplished with simple Auto Transformer and other means of soft starting. He has requested, and we are working on getting a university to do a test. Although, I have done this test myself many thousands of times, and it can easily be seen on a pump curve as well. He said a test by a university would help people to understand that the "cash incentives" should also be given for choosing the right pump and using a valve as control.

Over the years I have had many people try to go point by point and show me the error of my ways. I show them how to read a pump curve, and what I look for to choose the right pump. Then after a few minutes switching between a VFD and ATL with a valve on a working pump system, their eyes get big as saucers and the words "counter intuitive" become a regular part of their vocabulary. It really does work. It is just very difficult for people to understand when they already have incorrect preconceived notions about VFD's, and how valves "supposedly" burn energy.

I believe this is the correct thread and forum to discuss this. Harmonics are just one of many negative side effects of VFD control. I believe the best way to solve these problems is to not cause them in the first place. When a system can be controlled just as effectively and as efficiently without using a VFD, then all the problems associated with a VFD are never present to start with. Spend a little more time choosing the right pump, and a VFD cannot save any additional energy over simple valve controls.

Even on systems with existing and incorrectly sized pumps, a valve can simplify the controls and eliminate all the problems associated with VFD's. My customers in this position tell me that a little more energy use is a small price to pay, for the simplicity, dependability, and longevity of a pump system controlled by a valve rather than a VFD.

 

[Marke]

Hello valvecrazy

I am not sure that I am following your description.

the excess head produced is actually a free by-product

There is no free energy, the valve must have an increased pressure across it when it is restricting flow. It is essentially a hydraulic resistor and if you are operating into the same outlet such as an irrigator, it is going to restrict the flow by increasing the pressure across the valve and reducing the pressure at the irrigator. The pressure at the output of the pump will also increase, but this is dependent on the curve of the pump. If there is pressure drop across the valve, and flow through it, there must be a power loss in it.
I do not understand where this free byproduct comes from.

I certainly agree that the addition of a VFD to a system does introduce extra losses, typically in the order of 3% and anyone who uses a VFD on a fixed speed application is not doing their customers any favours, choosing the correct impeller for a fixed flow application and running at a fixed speed is sound engineering practice.

In a variable flow situation, it is difficult to optimize an impeller design/selection for all flow conditions. The use of a VFD effectively gives a similar result to an optimized impeller at each flow setting.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[gepman]

valvecrazy

Since you did not post a separate thread I thought that I would respond here.

#1

By knowing how to size and spec the right pumps, and by using mechanical means of varying the flow if necessary, there are NO energy savings by using VFDs.

This sentence does not quite make sense but I think that I know what you mean. For constant flow applications I don't think that anyone would argue with you that sizing the pump correctly and choosing a pump with a high efficiency at the expected pumping point is the best option. VFD's are used where variable loads or variable flows are expected. You state that one should use mechanical means for varying the flow if necessary. I assume that you would mean a valve in a liquid fluid system and a discharge damper in a fan system. Across every valve there is a flow (Q) and a pressure drop (H). The energy required to pump this flow across the valve is (Q*H)/(3960*%eff) where Q is flow in gpm, H is pressure drop across the valve, and %eff is the percent efficiency of the pump and motor combination. Therefore whenever you put a mechanical means to vary the flow it increases the amount of energy required to pump the fluid. It is true that if you increase the pressure drop across the valve the flow will decrease however the valve is still absorbing energy that does not need to be absorbed if you varied the flow with another method.

#2

Anytime you vary the speed of a pump, you are using more energy per gallon than if the pump were running at its designed BEP. Knowing this, how can anybody say varying the speed of a centrifugal pump with a VFD can save energy?

The first sentence is just not true AND it neglects the fact that when you vary the flow of a pump by "mechanical means", i.e. a valve, you are not running at the BEP and the energy per gallon will very likely be less with the pump that has had its speed varied than the pump that has its flow varied by a valve.

#3

Moving the sweet spot of the curve and maintaining maximum efficiency is just VFD propaganda, when the electric meter is still spinning at the same rate regardless. It is a common misconception that a VFD can slow a properly sized pump down enough to save energy.

The first part of the sentence may be true but you are not moving the "sweet spot" (BEP) with a VFD you are moving the curve. The "sweet spot" as you call it stays in the same relative position on the pump curve. When you vary the flow with a valve the "sweet spot" stays in the same position on the pump curve but the system curve moves out of the sweet spot. Therefore with a VFD you at least have a chance to remain at or near the BEP while with a valve you are almost guaranteed to move out of it (unless you were to the right of the sweet spot to begin with and then you wouldn't have sized you pump correctly as you advocate).

#4

Using your example of the variable flow situation, 1200 GPM, 800 GPM, or 400 GPM, the two curves follow each other so closely that all things considered, restricting with a valve reduces energy consumption as much as varying the speed.

In your first example at 100 gpm you are correct that the energy use is very close between a VFD and a valve throttled system. This was due to the low volume being pumped. Since energy is (as stated above) proportional to (Q*H)/(3960*%eff), when Q is low the energy being used is low. However look at your curves for 800 gpm, there is about a 15hp difference which is over a 17% savings.

#5

Of course a VFD works better with a pump that has a steep curve. This usually means that you have to oversize the pump to be able to (save energy?) by slowing it down with a drive.

A VFD works the same no matter what the slope of the pump curve. However the steeper pump curve will allow you to slow the pump down more in applications with high static head if that is what you mean. The slope of the pump curve does NOT mean that you have to oversize the pump or necessarily affect the efficiency of the pump. You choose the pump with the shape of the pump curve that you desire based upon your application. If you had an application where the static head varied considerably you would want a steeper pump curve so you at least still pump some water when the static head increased.

#6

Submersibles are about 80% of my business, and if the head requirement is 230', it will always be 230'. The well is not going to change from 230' deep to 120' deep just to help a VFD be more efficient. When the head is constant, as it is with nearly all pumping applications, it doesn't matter if it is submersible, centrifugal, or turbine, I can nearly always pick a pump that will save just as much energy by using a valve, as when using a VFD.

The water level in many wells will change depending on the weather. Some wells will easily change their standing water level by 100'. Then there is something call "drawdown". The well will have a replinishment rate usually measured in "ft of drop per 100 gpm". For a value of 10 this means that the well pumping water level drops 10 feet for every 100 gpm produced. Therefore at 1000 gpm the water level would drop 100' and at 500 gpm the water level would drop 50'. Therefore at 500 gpm output your static head would drop 50' from the 1000 gpm output. Also it is not true that almost all pumping applications are constant head. Most farmers (and other systems) really only care about the flow, not the head. The head is kept high so that the desired flow will occur in all situations. Another example of changing head requirements is a recent VFD application that I analyzed. The farmer had 200 acres of strawberries. When first planted he would sprinkler irrigate the strawberries which required a high flow and high head. Once established he would drip irrigate the strawberries which required a low flow and low head. A perfect VFD application. He got a $5,000 incentive from the local power utility to install the VFD based upon a 62,500 kWh savings for a year. Goulds (the largest pump manufacturer in the world and someone that makes their money in pumps, not VFD's) has introduced some rebranded VFD's (from ABB and A-B) with special software that really makes them quite clever and energy efficient. The link to this information is

http://www.ittmc.com/files/PumpSmart_Bulletin.pdf.

I highly recommend that you read it. You are correct in your thinking that the higher the percentage of static head/total head in your application, the less energy the VFD can save. However any time that you can slow the speed of the pump down instead of throttling you will save energy (neglecting VFD efficiency).

A true constant head system only exists in two scenarios, both of which do not occur very often. The first is if there is no flow. If there is no flow there is no friction loss and therefore head is constant although there is no need for a pump. The second is if there is no piping system. You don’t see that very often either. Actually if you had a pump between two tanks without piping there still is losses due to the contraction of the flow into the suction and expansion of the flow from the discharge. If it didn’t expand or contract then you would have to have a pipe on the suction and discharge. If you had a constant flow single circuit piping system it would be constant head. Then of course you would NOT need a valve to control flow. Any time you have a valve in place to control flow then it will not be a constant head system. Whatever pressure drop occurs across the valve could just as easily been achieved by reducing the speed of the pump to reduce the flow (and the pressure).

#7

Look at it this way. When the pump has been up-sized to work with a VFD, high flow of 1200 GPM at 100 HP is using .0833 HP per gallon. This is not even counting parasitic and other losses from the VFD. When slowed with a VFD to 100 GPM and using 40 HP, that is .4 HP per gallon produced. That means when the VFD is being used for low flow, it is actually burning .3167 extra horse power per gallon produced.

The pump is NOT upsized to work with a VFD. You can add the VFD to the pump that has been sized perfectly by you (or has already been installed by someone else). If you ever want to reduce the flow, you slow the pump down, you DON'T throttle the pump. I am attached some pump curves with system curves drawn on them. I think that the problem that you have in conceptualizing a VFD pumping system is that you do NOT see the system curve on the pump curve. The system curve is the head loss of the piping system as a function of flow. The head loss is proportional to the flow squared and therefore the system curve will be rising curve from left to right. The flow is ALWAYS were the pump curve and the system curve intersect. Closing a valve (throttling) changes the shape of the system curve and pushes it to the left. Slowing a pump down keeps the shape of the pump curve but reduces it size towards the origin of the pump curve (the same way that reducing the size of the impeller is shown on a pump curve).

I have taken the pump curve that you scanned and marked it up. I have assumed that the static head is 120’. If you try to state that the static head is 231’ then I will say that you could not have gotten 1200 gpm since 1200 gpm would have had at least 4.4’ of head loss in 200’ of pipe (out of the well). I will state again that the higher the percentage of static head/total head the less energy can be saved by using a VFD. Once static head/total head is greater than 80% it is very difficult to save any energy. Back to the example, at 100 gpm that you used previously, the pressure in the system would actually be about 121’ of head (see system curve) and therefore the VFD could be turning at about 72% speed or 2569 rpm. Assuming the same efficiency as that pumped at 1200 gpm (it won’t be but it will be more efficient than throttling) you would have the following horsepower (100*121)/(3960*.7) = 4.5 hp. Even if you assume the same efficiency as the throttled pump (about 13.88%) the horsepower used with the VFD is 22hp, about half that used with the valve.

If you don't understand my examples take a look at

http://www.maintenanceresources.com/referencelibrary/acdrives/vfd.htm.

In this article if you increased the static head the red line would just move upward but in its same relative position. If the pump curve was flatter you could not drop the speed of the VFD quite as much but you can still drop it more than you normally calculate because the shape of the system curve slopes downward to the left. You need to realize that only the static head is constant, the friction loss due to flow varies with the square of the flow.

The valve I am referring to only has 7 PSI friction loss at 1200 GPM, which is the only time a valve is actually burning energy. When a valve is used to further reduce the flow, the excess head produced is actually a free by-product, as throttling causes a reduction in power.

I think that you should re-examine this statement to see if you really want to stand by it. Let's examine it very closely. First let us assume a static head of 120', a little more than half of the 231' that you have at 1200 gpm. At full 1200 gpm your valve is burning (1200*7*2.31)/(3960*.7)= 7hp. At 100 gpm your valve has a pressure drop of 270-121=149’ (remember the system curve shows a total head of 121 feet at 100 gpm). Therefore the valve consumes (100*149)/(3960*.1388)=27.1hp. Yes your pump uses less energy than when it was pumping at 1200 gpm but it still wastes energy across the valve. Think of it this way, you could put a small turbine in place of the valve to produce the pressure drop and generate electricity.

I recommend for you the VFD savings calculator by ABB. It takes into account the static head, the efficiency of the VFD, the slope of the pump curve, and the present or proposed method of controlling flow (throttling valve, on/off, etc.). It then calculates the savings available for installing a VFD. You can find this software at

http://www.abb.us/product/ap/seitp322/24b03100d005c31ac1256e040043f4c1.aspx.

Play around with it and you can see how static head and pump curve shutoff head affect energy savings. I will vouch for the accuracy of the software.

I also see that you made a post since I started this humongous post. If you actually work with the either utility you know that they have a fleet of over 40 pump testers who test efficiencies of pumps, that these pump testers recommend VFD’s in certain specific cases, and that what you describe does not even follow the SPC (Standard Performance Contract) energy savings software that both utilities REQUIRE to be used for energy savings calculations (you can download it at

http://www.aesc-inc.com/download/spc/).

I don’t expect this post to sway you much but I would be happy to meet you at either PG&E’s Pacific Energy Center in San Francisco, SCE’s CTAC in Irwindale, or SCE’s AgTAC in Tulare if you need to see how VFD's work in pumping systems. The latter would probably be best so we could do a pump test right on the site (they already have it set up to prove everything that you say isn’t so).

 

[gepman]

Marke
Here is something that has been investigated in California as a replacement for VFD's. It is still a VSD (Variable Speed Drive) so it won't be of any interest to valvecrazy but since it is non-electrical it won't create any harmonics. I don't see how it could be used on a submersible pump but it obviously will work for a horizontal pump and I think that it could be adapted for a vertical mount turbine or lineshaft pump.

 

[Marke]

Hello gepman

As I read that paper, it appears to be a form of magnetic coupling where the input torque and output torque are equal. In a sense, this is similar to using a valve in that it is a series element introducing losses.
If the output shaft was spinning at 75% speed relative to the input shaft, and there is equal input and output torque, then there will be 25% slip losses in the coupling. If I am wrong, and the output toque increases with the slip then this could be worthwhile looking at.

From all the information that I have gleaned so far, it would appear that using different supply transformers on the individual pumps such that there is a balance of transformer load on two groups of transformers, one having an additional phase shift will result in cancellation of the 5th and 7th harmonics. This is good because it means that if new installations are all done with the additional phase shift transformers, the net result will be a reduction in the existing harmonics.

Alternative considerations are fitting filters to all drives, adding active filters to the supply, converting existing drives to active front end etc. All of which are more expensive and disruptive to existing installations.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[itsmoked]

Wow Gepman!
Thanks for making the effort.
Most interesting.


Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

This is easily one of the most valuable threads in the Motor and Controls forum. Even if the true truth always lies just out of (at least my) reach, it has been a refreshing experience to see dedicated experts discuss these matters so in-depth. I am not to judge, even if I am more in favour of sound physical reasoning than gut-feeling, but I think that gepman backs up his claims more solidly than his opponent.

The magnetic coupler, that was thrown in late in this discussion, is a very good illustration that many engineers and economists tend to forget that slip always is equivalent to losses. As is breaking or any attempt to reduce a movement or a flow without reducing the input power. It is only by reducing input power, while maintaining design goals, that any energy can be won.

The main techniques for reducing input power is:
1. To select correctly sized components (Valvecrazy is absolutely correct there)
2. To select the right operating speed. Be it with the aid of gears, pulleys, using a VFD or any other LOSSLESS means.
3. To select piping and armatures with low resistance to flow.
4. To select motors with high efficiency - and supply them with the right voltage and frequency for the task at hand. Which may be variable.
5. To avoid any device that introduces losses. Like valves, dampers, magnetic couplings or introducing extra slip in motors (the WRIM does the same thing as the magnetic coupler - it increases slip and burns energy).

As a last suggestion, what about hydraulic drives? I have worked with Hägglund drives. They are sometimes superior to any other drive system - although I do not think that they fit very well in Mark's pump systems. But they certainly do not produce any harmonics or EMI of any frequency.


Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[DickDV]

That magnetic coupling is nothing more than a clever repackaging of the old Eddy Current Clutch, long ago discredited except on affinity rule loads like centrifugal pumps and centrifugal fans.

Since they do nothing to reduce motor excitation at light loads, you can reasonably expect that they will save some energy but not as much as a properly spec'ed VFD/motor system.

I've seen similar magnetic couplings marketed in HVAC for fan loads and, other than the harmonics issues, have little to make them attractive, in my opinion.

 

[ScottyUK]

I have found the elusive eddy current drive thread which both Dick and I looked for a while ago. Seems the 'search' function really is a 'hide' function!

The thread in which we were looking for the missing thread: thread237-107895

and the missing thread which disappeared for a while: thread237-98040


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Valvecrazy]

The systems I work with have a much flatter system curve. The well is 200' to water. Static pulls down very little and very quickly, so the feet of lift is constant. I am also a long time licensed well driller. The drip system on the strawberry patch needs 10 PSI constant. The underground lines are large enough to have only a couple of PSI of friction loss at 1200 GPM, or max flow. The variable is that we irrigate the field in thirds. Sometimes at 400, sometimes 800, and other times 1200 GPM. So the head is always 230', I just have to choose a pump that has good brake horsepower characteristics for variable flows. If I spend my time choosing the correct pump, the horse power difference is similar to the first curve I posted. I do not argue that there are a few points along that curve where the VFD shows a little lower horse power requirements. You say 15 HP at 800 GPM, it looks more like 12 HP to me. Add back in the 3% losses from a drive, which I think is low, and you are saving 9 HP or 9% at that flow rate. At 100 GPM is only 4 HP, with losses is really 1 HP in savings with a VFD. At 1200 GPM the VFD is actually wasting at least 3% to drive losses compared to ATL controls. Average all this up, depending on the run time at the various flow rates, and you might average 4% energy savings with VFD. I cannot argue that.

Now will that 4% ever pay for the added expense of the Drive itself, before the electronics or drive must be replaced? Is that 4% worth the problems of harmonics, which is actually dirty power and causes everything connected to the grid to be less efficient? Then we could talk about bearing currents, resonance frequencies, voltage spikes and other things. One of the biggest problems is longevity. Motors would not have tags on them bragging about "Voltage Spike Resistant Inverter Ready Windings" if the VFD where not harder on motors than running Across The Line. So now motors don't last as long and drives don't last as long as equipment running ATL. How much energy does it take to mine, manufacture, transport, install, and recycle the equipment that did not last at long as it should have?

Then if energy efficiency is that important or the system curve varies as dramatically as yours, I would use a multiple pump set up. It is very easy to stagger the pressure up or down as needed, and switch between different pumps when using valves for control. If I needed a 100 HP maximum load, the most efficient way would be to use a 60 HP, 30 HP, and a 10 HP. You can't get anymore efficient than a 10 HP pumping 100 GPM, or a 10 HP and a 30 HP running together to pump 400 GPM. 800 GPM would take a 60 and 30 running together, while max of 1200 GPM would require all three pumps adding up to 100 HP.

A Flow Based Pressure Management system would use two pumps to produce different pressures. When the system curve varies from 120' at 400 GPM to 230' at 1200 GPM I would have a 100 HP doing the high flow high pressure part of the curve. Then a 20 HP could do the low pressure low flow part much more efficiently.

I have read the PumpSmart thing and have written AB about their so called calculator. After all they are trying to sell this product. The main argument I have is that people rarely give the centrifugal pump credit for it's "magical", "counter intuitive" properties, that causes it to reduce it's power consumption when flow is throttled. If you know how to take advantage of this, you can stay within a percent or two of what a VFD can do, without all the complications that a VFD adds to the scenario. For goodness sake, if you don't think a VFD adds complications, just look at this thread!!!!!

 

[Marke]

Hello valvecrazy

Yes, agreed there are situations where the correct selection of pump and impeller can give admirable results and a VFD adds nothing. There are many of these installations out there where a VFD has been used, perhaps with a less optimized pump.

There are however, situations where the applications can not be adequately catered for using a single fixed speed pump.

A Flow Based Pressure Management system would use two pumps to produce different pressures. When the system curve varies from 120' at 400 GPM to 230' at 1200 GPM I would have a 100 HP doing the high flow high pressure part of the curve. Then a 20 HP could do the low pressure low flow part much more efficiently.

A multi pump installation is one solution and a VFD with one pump is another. I believe that there is merit in both solutions with the VFD effectively scaling the size of the impeller in performance.

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[nightfox1925]

I appreciate the efforts and expertise everyone in this thread had given. I ma very much interested on this thread and I keep on reading all the technical inputs several times. Pumps and undertanding pump curves is way out of my league but I do have the interest to learn how to undertand them...so at least i could get a clearer picture myself on this thread which i find very informative. Is there any link or reading that would give me some simolified way of reading the pump curves? Thanks

GO PLACIDLY, AMIDST THE NOISE AND HASTE-Desiderata